• Mitchell Boehnke, David Mirsky, Nicholas Stence, Rachel M Stanley, Daniel M Lindberg, and ExSTRA investigators.
• Department of Diagnostic Radiology, Children's Hospital Colorado, 12631 East 17th Ave., Mail Stop 8200, Aurora, CO, 80045, USA. Mitchell.boehnke@ucdenver.edu.
• Pediatr Radiol. 2018 Aug 1; 48 (8): 1123-1129.

BackgroundStudies evaluating small patient cohorts have found a high, but variable, rate of occult head injury in children <2 years old with concern for physical abuse. The American College of Radiology (ACR) recommends clinicians have a low threshold to obtain neuroimaging in these patients.ObjectivesOur aim was to determine the prevalence of occult head injury in a large patient cohort with suspected physical abuse using similar selection criteria from previous studies. Additionally, we evaluated proposed risk factors for associations with occult head injury.Materials And MethodsThis was a retrospective, secondary analysis of data collected by an observational study of 20 U.S. child abuse teams that evaluated children who underwent subspecialty evaluation for concern of abuse. We evaluated children <2 years old and excluded those with abnormal mental status, bulging fontanelle, seizure, respiratory arrest, underlying neurological condition, focal neurological deficit or scalp injury.ResultsOne thousand one hundred forty-three subjects met inclusion criteria and 62.5% (714) underwent neuroimaging with either head computed tomography or magnetic resonance imaging. We found an occult head injury prevalence of 19.7% (141). Subjects with emesis (odds ratio [OR] 3.5, 95% confidence interval [CI] 1.8-6.8), macrocephaly (OR 8.5, 95% CI 3.7-20.2), and loss of consciousness (OR 5.1, 95% CI 1.2-22.9) had higher odds of occult head injury.ConclusionOur results show a high prevalence of occult head injury in patients <2 years old with suspected physical abuse. Our data support the ACR recommendation that clinicians should have a low threshold to perform neuroimaging in patients <2 years of age.

26 keywords...

hide…